data analysis hw

Today we will explore graphs useful for displaying quantitative data.  Recall that quantitative data is date that deals with numbers rather than categories. In the titanic file, some of the columns are qualitative, such as pclass and survived, others are quantitative, such as age.

Histograms

1. Load and attach the data and plot the histograms for age column, then insert the graph in this file:

#plot a histogram

hist(age)

hist(age,breaks=c(0,10,20,30,40,50,60,70,80)) #changes the bin sizes

2. Estimate based on the graph you found how many passengers were under age 10.

3. Create a new histogram in which survived = 0.

#Choose some observations from a variable by another variable

L <- survived==0

died.age <- age[L]

hist(died.age,breaks=c(0,10,20,30,40,50,60,70,80))

Insert the graph here:

4. Create another histogram, in which survived = 1

survived.age=age[!L]

hist(survived.age,breaks=c(0,10,20,30,40,50,60,70,80))

Insert it here:

5. Compare the histograms of the ages of those that survived vs. those that died. Especially look at the number of children. What do these graphs imply?

Dot plots

6. Load the file babyboom.csv. This data set comes from an article in the Journal of Statistics Education, v7n3. It contains the time of birth, sex (1=girl, 2=boy), and birth weight for 44 babies born in a 24 hour period at a particular hospital. Create dot plots for birth weight.

#load the data

babyboom<-read.csv(file.choose(),header=TRUE)

attach(babyboom)

#dot plot for birth weight

stripchart(Weight,method="stack",pch=19) #try leaving off the “method” and “pch” and see what happens

# plot(Weight) Try this and compare the graphs.

Insert the graph below.

 

A dotplot displays a dot for each data point. If two data points are the same, then two dots will be stacked on top of each other.

7. Create dot plots again, but this time group the data by sex.

stripchart(Weight~Sex, method="stack", pch=19)

Insert your graph below. Does it look like there is a big difference in birth weight for boys vs. girls?

 

Comparing the two graphs  

8. Clearly there are situations in which both histograms and dot plots can be used.  When is it better to use a histogram, and when is it better to use a dot plot?

Percentiles and Boxplots

9. #summarize birth weight

summary(Weight)

Look at the values reported, and list the ones requested here:

Min:

Q1:

Med:

Q3:

Max:  

These five numbers (Minimum, First Quartile, Median, Third Quartile, and Maximum) divide the data set into four pieces, with 25% of the data between each neighbor pair. Together, they are referred to as the five number summary

10. #create boxplots

boxplot(Weight)

Insert the graph below.

The boxplot is no more than a visualization of the five number summary. The location of the bottom of the graph matches the minimum, the location of the bottom of the box is the first quartile, the middle of the box is at the median, the top of the box is at the third quartile, and the top of the graph is at the maximum. Occasionally, as in this case, you will see a boxplot that has extra dots outside the “whiskers.” These dots represent values that have been deemed unusual compared to the rest of the data set. These values are called outliers.

11. Create boxplots again, this time grouped by sex. (#boxplots grouped by sex

boxplot(Weight~Sex)) Also find the five number summaries of birth weights grouped by sex.

Boys:	Girls:
Min:	Min:
Q1:	Q1:
Med:	Med:
Q3:	Q3:
Max:	Max:

12. Describe any difference you observe between the boxplots for the boys' weights and the girls' weights.

13. Back to the Titanic. Create appropriate graphs that show the ages of the survivors, grouped by passenger class. Comment on what the graphs show in terms of what happened that night.